基于视频跟踪轨迹的全过程路侧停车行为检测与识别技术

针对路侧停车行为检测与识别问题，结合经典KLT运动角点检测方法与实时压缩跟踪方法，提出一种基于视频跟踪轨迹的全过程停车行为识别技术。首先利用KLT方法检测视频中的运动角点，确定运动车辆位置，再利用压缩跟踪方法，提取视频序列中运动车辆完整的运动轨迹，较为准确地描述路侧停车的动态过程。最后，利用提取的运动特征在真实的路侧停车视频上进行实验验证。实验结果表明，本文提出的停车行为检测与识别方法是可行有效的。 

A Roadside Parking Detection and Recognition Method Based on Video Motion Trail

An effective detection and recognition method is proposed for the roadside parking problem by combining the KLT corners detection approach and real-time compressive tracking (CT) approach. First, KLT approach is used to detect corner in video for extracting the area of moving car. And then the CT approach is adopted to acquire the motion trail for each parking action to describe the dynamic process of moving cars completely. The proposed method is tested on the real video data for evaluation. The results demonstrate that this method is feasible and effective.